Installation issues with latest replicated on k8s

Packaging an application
1

Hi,

I have created my application jarvisk8s under CA company namespace. Generated a license. Fired off the install command based on what I see in releases (unstable).

Installation went fine for k8s & replicated. And then it suggests that I open admin console & reset password. And that’s where things started going wrong.

  1. It throws an LDAP error during password reset. even if i choose anonymous login, error is the same.
  2. once this error hits, if reopen the admin console URL, it will never respond. - https://IP:8800/create-password
  3. after a while i see that my services & pods do show up on k8s but all pods failed with ImagePullBackOff error and my images are in replicated docker registry.

I am currently blocked on this issue & its consistently reproducible on ubunutu & centos (onprem or aws).
Appreciate your help.

Thanks & Regards
Anil

2

There should be some errors in the replicated container logs, which you can get by running `kubectl logs replicated.

You can also check events for the pods failing with the pull back off error.

3

replicated-login-error
replicated-login-error1409×1206 102 KB

4

I dont see any pods in the NS replicated-9ec6675d3b5b4181519356c5d8f84bc3
in the default NS, there are 2 replicated pods

kubectl get pods | grep replicated
replicated-premkit-6d8dd87586-v5p8n   2/2       Running             0          2d
replicated-statsd-7454dd4f6d-6dzdn    0/2       ContainerCreating   0          2d
5

kubectl describe pod replicated-statsd-7454dd4f6d-6dzdn

 Events:
  Type     Reason       Age                 From                      Message
  ----     ------       ----                ----                      -------
  Warning  FailedMount  6m (x1273 over 2d)  kubelet, ip-172-31-3-167  Unable to mount volumes for pod 
"replicated-statsd-7454dd4f6d-6dzdn_default(2a80314d-95de-11e8-8ac7-0a14515dc2b4)": timeout expired waiting for volumes to attach/mount for pod "default"/"replicated-statsd-7454dd4f6d-6dzdn". list of unattached/unmounted volumes=[replicated-statsd-graphite-storage]
6

I just did a fresh install on centos7.4

at the end of it, uploaded license file & at login, simply chose anonymous for admin login. Got the same error. This clearly looks like a bug to me.

Error response from LDAP server
Error: Request has been terminated Possible causes: the network is offline, Origin is not allowed by 
Access-Control-Allow-Origin, the page is being unloaded, etc.
7

so the problem was namespace entries in my k8s yamls. once i removed them all in my definitions, it just worked. thanks.

8

and the reason is that replicated uses “default” namespace for itself and creates a random “replicated-someUUID” NS for apps.